The SIMULINK- models of multibit Delta-Sigma ADC and Delta-Sigma DAC all have a non-ideal D/A-converter.
The DAC mismatch noise will be shaped according to a general Data Weighted Averaging (DWA) transfer function 1±z^(-D), where D is a positive integer.
The DWA element selection logic calculates a selection vector. The selection vector chooses, which of the unit DACs will be used according to DAC input.
The DWA block is constructed using mainly fixed-point blocks.
Models for Incremental DWA (IDWA) were also added.
This package also includes a spurious tone estimator: the spurs can be estimated using the DS input signal and the DNL error shape.
This model and the related paper (mentioned previously) were also a part of my Ph.D. thesis:
The mismatch noise estimator was improved ad added to submission #29522: "Delta Sigma converter spurious tone predictor".
Both DWA mismatch noise model and DSM quantization noise models can be found in #29522.
More information about this model:
Neitola, M., Rahkonen, T.: A Generalized Data-Weighted Averaging Algorithm. Circuits and Systems II: Express Briefs, IEEE Transactions on, Volume: 57 , Issue: 2, Publication Year: 2010 , Page(s): 115 - 119
Available on IEEE Xplore
A bugfix: the spur estimator now supports all DWA modes (i.e. noise shaping transfer functions).
Feature update: spurious tone estimator.
Added three simulation models for Delta-Sigma D/A converters.
Major improvement. The DWA-DAC is now more realistic. Also added a model that uses incremental DWA.